@Nawoda Dharmabandu Based on my experience with the service the performance of the API calls to extract from files depends on the model that is trained for custom model scenarios and for prebuilt models provided by Azure the response is also good. The official documentation does not provide limits to the requests/second or throttling thresholds as it is restricted by the input requirements and other limitations and best practices. Since all the API calls to extract forms are async the POST requests if accepted provide a 202 response and the processed result is returned via GET results API. If the result is available/unavailable with GET an appropriate status message will be indicated in the response. I hope this helps.
Azure Cognitive Service Form Recognizer
Nawoda Dharmabandu
1
Reputation point
I'm trying to identify the best services to extract pdf/image text to be used in a project. And I am pretty interested in the Cognitive Service Form Recognizer. So I very much appreciate it if someone can help me through the following.
- What will be the processing time for one pdf when using Azure Cognitive Service Form Recognizer and Power Automate to Extract PDF Text?
- What will be the request throttling thresholds?
1 answer
Sort by: Most helpful
-
romungi-MSFT 42,771 Reputation points Microsoft Employee
2021-08-03T10:24:51.487+00:00